1. Field of the Invention
The present invention relates to a document processor for processing a language using Chinese characters, i.e., so-called "kanji" such as in the Japanese and Chinese language, and more particularly to a document processor adapted to retrieve a desired kanji character from its corresponding phonetic "kana" symbols.
2. Discussion of the Related Art
Japanese word processors as well as work stations and computers that are capable of processing the Japanese language are in widespread use for editing Japanese sentences. Generally, such an apparatus includes a kana to kanji conversion system in which Japanese phonetic kana symbols (hereinafter referred to as the reading) are input to the processor which then generates corresponding Japanese sentences composed of "hiragana", "kanji", "katakana" and the like.
When inputting kana symbols for conversion to kanji characters by the kana to kanji conversion method, particularly when inputting a name of a person or place which may use kanji characters that are not frequently used, corresponding kanji characters may not be available in a dictionary memory used by the conversion method. This is also true when inputting the full name of a person or place. Therefore, in these cases, kana to kanji conversion is impossible. Accordingly, in such a case it is the common practice to effect single-kanji conversion in which kanji are input one by one. For instance, if the Japanese name " .sup.3 " (pronounced "Kouichi" in English) is input by using kana " .sup.4 " as a phonetic symbol, it is impossible to convert the Japanese phonetic kana symbols into a corresponding kanji name using an ordinary dictionary. Hence, the name " .sup.3 " is divided into two characters " .sup.3 " and " .sup.3 " so that it can be input sequentially, thereby completing the name " .sup.3 ".
With such a conventional document processor, in the case of a reading such as " .sup.2, " numerous kanji may be identified in an ordinary dictionary. In cases where such homophonic kanji exist in large numbers, if they are displayed and selected one by one, the operation until a desired set of characters is obtained becomes complicated. In addition, the time required in the selecting operation is disadvantageously long.
Accordingly, in some devices a measure is adopted in which the sequence of displaying the homophonic kanji is set as the sequence of frequency of their appearance. Although it is possible with such a technique to efficiently select popular kanji, if an attempt is made to select a kanji of a low use frequency, more time is required for the selection process.